Objective-We tested the hypothesis that deficiency of cellular glutathione peroxidase (GPx-1) enhances susceptibility to endothelial dysfunction in mice with moderate hyperhomocysteinemia. Methods and Results-Mice that were wild type (Gpx1 ϩ/ϩ ), heterozygous (Gpx1 ϩ/Ϫ ), or homozygous (Gpx1 Ϫ/Ϫ ) for the mutated Gpx1 allele were fed a control diet or a high-methionine diet for 17 weeks. Plasma total homocysteine was elevated in mice on the high-methionine diet compared with mice on the control diet (23Ϯ3 versus 6Ϯ0.3 mol/L, respectively; PϽ0.001) and was not influenced by Gpx1 genotype. In mice fed the control diet, maximal relaxation of the aorta in response to the endothelium-dependent dilator acetylcholine (10 Ϫ5 mol/L) was similar in Gpx1 ϩ/ϩ , Gpx1 ϩ/Ϫ , and Gpx1 Ϫ/Ϫ mice, but relaxation to lower concentrations of acetylcholine was selectively impaired in Gpx1 Ϫ/Ϫ mice (PϽ0.05 versus Gpx1 ϩ/ϩ mice). In mice fed the high-methionine diet, relaxation to low and high concentrations of acetylcholine was impaired in Gpx1 PϽ0.05). No differences in vasorelaxation to nitroprusside or papaverine were observed between Gpx1 ϩ/ϩ and Gpx1 Ϫ/Ϫ mice fed either diet. Dihydroethidium fluorescence, a marker of superoxide, was elevated in Gpx1 Ϫ/Ϫ mice fed the high-methionine diet (PϽ0.05 versus Gpx1 ϩ/ϩ mice fed the control diet). Conclusions-These findings demonstrate that deficiency of GPx-1 exacerbates endothelial dysfunction in hyperhomocysteinemic mice and provide support for the hypothesis that hyperhomocysteinemia contributes to endothelial dysfunction through a peroxide-dependent oxidative mechanism. Key Words: endothelium Ⅲ homocysteine Ⅲ nitric oxide Ⅲ peroxide H yperhomocysteinemia is an emerging risk factor for cardiovascular events and venous thrombosis, 1,2 but the mechanisms responsible for the vascular pathology of hyperhomocysteinemia are still incompletely understood. Like many other cardiovascular risk factors, hyperhomocysteinemia produces endothelial dysfunction, which is possibly due to oxidative inactivation of endothelium-derived NO. 3,4 The oxidative stress hypothesis 4 is supported by the observation that auto-oxidation of homocysteine in vitro generates reactive oxygen species (ROS), including hydrogen peroxide and superoxide, and promotes oxidation of LDL. 5-7 Treatment of cultured endothelial cells with homocysteine decreases bioavailable NO 8,9 and produces cytotoxicity mediated by hydrogen peroxide. 10 Homocysteine also indirectly contributes to oxidative stress by inhibiting the expression of antioxidant enzymes, such as cellular glutathione peroxidase (GPx-1), an effect that may sensitize one to the toxic effects of homocysteine-derived hydrogen peroxides and lipid peroxides. 8,11 However, the role of oxidative stress in the vascular dysfunction of hyperhomocysteinemia in vivo is less clear, and its clinical importance has been questioned. 12 Attempts to demonstrate elevated levels of oxidation products in humans with hyperhomocysteinemia have produced conflicting results. [13][14][15][16] R...